Color cast with PhotometricColorCalibration

[astroman2]

I am using PCC to correct color on a linear frame from a segment near M42.  PCC plate solves correctly and seems to run correctly, but creates a strong red cast.  If I run PCC a second time, the color gets even redder.

If I use ColorCalibration to correct the frame color, I get expected results and removal of a color cast.

I have attached the files in the drop box links below.  Can anyone advise what I am doing wrong?

https://www.dropbox.com/s/cs63eydwbcug1b3/pcc_test.xpsm?dl=0
https://www.dropbox.com/s/hs56fboxrwsg16d/M42-1_300secs_32%20frames_integration_beforePCC.xisf?dl=0
https://www.dropbox.com/s/1sg06xlto3whn03/M42-1_300secs_32%20frames_integration_afterPCC.xisf?dl=0

Thanks

Scott

[Alejandro Tombolini]

Hi Scott, I did color calibration in your image using PCC. You can follow the steps here. Hope this helps.

Saludos, Alejandro.

[Juan Conejero]

Hi Alejandro,

Excellent tutorial and results. A nice practical example of PCC where you achieved perfect white balancing.

[GJL]

Alejandro, you´re writing in the tutorial:

"Finally, this tool allows to perform background neutralization of the image in the same process. For this, choose a preview that better represent the background of the image, drag it into the Region of interest of the tool and set the Upper limit value with a value lower than the maximun value of pixel in the preview."

But you don't set the upper limit value lower but higher to 0.007, i.e. higher than 0.0069...
What is true?

Gerhard

[Alejandro Tombolini]

Hi Gerhard, thank you for the observation, you are right!

I have corrected it: set the Upper limit value with a value higher than the maximum value of pixel in the preview

Juan, thank you! 

Saludos, Alejandro.

[astroman2]

Alejandro,

Thanks for the demonstration.  I can match your result.

It seems that the pcc_test icon I generated has an error in the image coordinates.  If you run the test without forcing plate solving, pcc appears to run and solve, but generates an incorrect color correction (and the photometry graphs have a negative slope).  If I force plate solving, the color correction works perfectly.

Thanks

Scott